A temporal action proposal system combining C3D features, a self-attention pyramid network, and boundary refinement reported 71.38% test AUC and second place in the ActivityNet 2019 challenge.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Relation-Aware Pyramid Network (RapNet) for temporal action proposal
A temporal action proposal system combining C3D features, a self-attention pyramid network, and boundary refinement reported 71.38% test AUC and second place in the ActivityNet 2019 challenge.